a b s t r a c tThe study sought to determine the prevalence and predictors of being at risk of an alcohol-exposed pregnancy (AEP) among women of child-bearing age in an urban and rural location in South Africa. We conducted a cross-sectional household survey of 1018 women aged 18-44 years in one urban (n ¼ 606) and one rural (n ¼ 412) site. The women were interviewed using a structured questionnaire. We defined the primary dependent variable, being at risk of having an AEP, as current alcohol use, not being pregnant, being fertile, and no effective use of contraceptives. The independent variables included demographic, substance use, health perceptions, psycho-social, and partner characteristics. The rural women (21.84%) were more likely than their urban counterparts (11.22%) to be at risk of an AEP. In multiple logistic regression analyses, significant predictors of being in the ''at risk'' group for the urban women were (a) being 'white' as opposed to 'black/African', and being 'coloured' as opposed to 'black/African'; and (b) current smoking. For the rural women, significant risk factors were (a) current smoking and (b) early onset of alcohol use. The significant protective factors were (a) education; (b) knowledge about Fetal Alcohol Syndrome; (c) parity. Use of stricter alcohol use criteria (i.e., three or more drinks and five or more drinks per sitting) in the definition of risk of an AEP yielded slightly different patterns of significant predictors. The results revealed high levels of risk of an alcohol-exposed pregnancy, especially amongst the rural women, and a need for location-specific prevention programmes. The high burden of AEP in South Africa calls for the establishment of national AEP prevention strategies and programmes as a matter of urgency.